Data-driven Consensus Protocol for Regulating Multi-Stack PEMFC Systems

This study proposes a novel data-driven consensus protocol for regulating the required oxygen excess ratio (OER) of multi-stack Proton Exchange Membrane Fuel Cell (PEMFC) systems, without requiring real-time state access or prior knowledge of the system dynamics. The control scheme incorporates a distributed state observer that predicts the states of each fuel cell by leveraging relative information from neighboring fuel cells. A consensus algorithm is then employed to enable each follower fuel cell to track the OER signal of the leader system. To optimize the consensus feedback gain, a direct data-driven optimal controller is designed, relying solely on leader system data. Furthermore, sufficient conditions for stability are derived from Lyapunov theorem and datadependent linear matrix inequalities solutions. The effectiveness of the proposed approach, along with the theoretical findings, are validated through numerical simulations, demonstrating robust performance under load variation and network uncertainties, including Deniel of Service and False Data injection attacks, with accurate tracking of the leader's outputs across all PEMFC stacks.